• Forests harbour large spatiotemporal heterogeneity in canopy structure. This variation drives the microclimate and light availability at the forest floor. So far, we do not know how light availability and sub-canopy temperature interactively mediate the impact of macroclimate warming on understorey communities.
• We therefore assessed the functional response of understorey plant communities to warming and light addition in a full factorial experiment installed in temperate deciduous forests across Europe along natural microclimate, light and macroclimate gradients. Furthermore, we related these functional responses to the species’ life-history syndromes and thermal niches.
• We found no significant community responses to the warming treatment. The light treatment, however, had a stronger impact on communities, mainly due to responses by fast-colonizing generalists and not by slow-colonizing forest specialists. The forest structure strongly mediated the response to light addition and also had a clear impact on functional traits and total plant cover.
• The effects of short-term experimental warming were small and suggest a time-lag in the response of understorey species to climate change. Canopy disturbance, for instance due to drought, pests or logging, has a strong and immediate impact and particularly favours generalists in the understorey in structurally complex forests.

Context Evidence for effects of habitat loss and fragmentation on the viability of temperate forest herb populations in agricultural landscapes is so far based on population genetic studies of single species in single landscapes. However, forest herbs differ in their life histories, and landscapes have different environments, structures and histories, making generalizations difficult.

Objectives We compare the response of three slow-colonizing forest herbs to habitat loss and fragmentation and set this in relation to differences in life-history traits, in particular their mating system and associated pollinators.

Methods We analysed the herbs' landscape-scale population genetic structure based on microsatellite markers from replicate forest fragments across seven European agricultural landscapes.

Results All species responded to reductions in population size with a decrease in allelic richness and an increase in genetic differentiation among populations. Genetic differentiation also increased with enhanced spatial isolation. In addition, each species showed unique responses. Heterozygosity in the self-compatible Oxalis acetosella was reduced in smaller populations. The genetic diversity of Anemone nemorosa, whose main pollinators are less mobile, decreased with increasing spatial isolation, but not that of the bumblebee-pollinated Polygonatum multiflorum.

Conclusions Our study indicates that habitat loss and fragmentation compromise the long-term viability of slow-colonizing forest herbs despite their ability to persist for many decades by clonal propagation. The distinct responses of the three species studied within the same landscapes confirm the need of multi-species approaches. The mobility of associated pollinators should be considered an important determinant of forest herbs' sensitivity to habitat loss and fragmentation.

Hedgerows have the potential to facilitate the persistence and migration of species across landscapes, mostly due to benign microclimatic conditions. This thermal buffering function may become even more important in the future for species migration under climate change. Unfortunately, there is a lack of empirical studies quantifying the microclimate of hedgerows, particularly at broad geographical scales. Here we monitored sub-canopy temperatures using 168 miniature temperature sensors distributed along woodland-hedgerow transects, and spanning a 1600-km macroclimatic gradient across Europe. First, we assessed the variation in the temperature offset (that is, the difference between sub-canopy and corresponding macroclimate temperatures) for minimum, mean and maximum temperatures along the woodland-hedgerow transects. Next, we linked the observed patterns to macroclimate temperatures as well as canopy structure, overstorey composition and hedgerow characteristics. The sub-canopy versus macroclimate temperature offset was on average 0.10 degrees C lower in hedgerows than in woodlands. Minimum winter temperatures were consistently lower by 0.10 degrees C in hedgerows than in woodlands, while maximum summer temperatures were 0.80 degrees C higher, albeit mainly around the woodland-hedgerow ecotone. The temperature offset was often negatively correlated with macroclimate temperatures. The slope of this relationship was lower for maximum temperatures in hedgerows than in woodlands. During summer, canopy cover, tree height and hedgerow width had strong cooling effects on maximum mid-day temperatures in hedgerows. The effects of shrub height, shrub cover and shade-casting ability, however, were not significant. To our knowledge, this is the first study to quantify hedgerow microclimates along a continental-scale environmental gradient. We show that hedgerows are less efficient thermal insulators than woodlands, especially at high ambient temperatures (e.g. on warm summer days). This knowledge will not only result in better predictions of species distribution across fragmented landscapes, but will also help to elaborate efficient strategies for biodiversity conservation and landscape planning.

Questions: Does the influence of forest edges on plant species richness and composition depend on forest management? Do forest specialists and generalists show contrasting patterns?

Location: Mesic, deciduous forests across Europe.

Methods: Vegetation surveys were performed in forests with three management types (unthinned, thinned 5-10 years ago and recently thinned) along a macroclimatic gradient from Italy to Norway. In each of 45 forests, we established five vegetation plots along a south-facing edge-to-interior gradient (n = 225). Forest specialist, generalist and total species richness, as well as evenness and proportion of specialists, were tested as a function of the management type and distance to the edge while accounting for several environmental variables (e.g. landscape composition and soil characteristics). Magnitude and distance of edge influence were estimated for species richness per management type.

Results: Greatest total species richness was found in thinned forests. Edge influence on generalist plant species richness was contingent on the management type, with the smallest decrease in species richness from the edge-to-interior in unthinned forests. In addition, generalist richness increased with the proportion of forests in the surrounding landscape and decreased in forests dominated by tree species that cast more shade. Forest specialist species richness, however, was not affected by management type or distance to the edge, and only increased with pH and increasing proportion of forests in the landscape.

Conclusions: Forest thinning affects the plant community composition along edge-to-interior transects of European forests, with richness of forest specialists and generalists responding differently. Therefore, future studies should take the forest management into account when interpreting edge-to-interior because both modify the microclimate, soil processes and deposition of polluting aerosols. This interaction is key to predict the effects of global change on forest plants in landscapes characterized by the mosaic of forest patches and agricultural land that is typical for Europe.

Linear landscape elements such as hedgerows and road verges have the potential to mitigate the adverse effects of habitat fragmentation and climate change on species, for instance, by serving as a refuge habitat or by improving functional connectivity across the landscape. However, so far this hypothesis has not been evaluated at large spatial scales, preventing us from making generalized conclusions about their efficacy and implementation in conservation policies. Here, we assessed plant diversity patterns in 336 vegetation plots distributed along hedgerows and road verges, spanning a macro-environmental gradient across temperate Europe. We compared herb-layer species richness and composition in these linear elements with the respective seed-source (core) habitats, that is, semi-natural forests and grasslands. Next, we assessed how these differences related to several environmental drivers acting either locally, at the landscape level or along the studied macro-ecological gradient. Across all regions, about 55% of the plant species were shared between forests and hedgerows, and 52% between grasslands and road verges. Habitat-specialist richness was 11% lower in the linear habitats than in the core habitats, while generalist richness was 14% higher. The difference in floristic composition between both habitat types was mainly due to species turnover, and not nestedness. Most notably, forest-specialist richness in hedgerows responded positively to tree cover, tree height and the proportion of forests in the surrounding landscape, while generalist richness was negatively affected by tree height and buffering effect of trees on subcanopy temperatures. Grassland and road verge diversity was mainly influenced by soil properties, with positive effects of basic cation levels on the number of specialists and those of bioavailable soil phosphorus on generalist diversity. Synthesis and applications. We demonstrate that linear landscape elements provide a potential habitat for plant species across Europe, including slow-colonizing specialists. Additionally, our results stress the possibility for land managers to modify local habitat features (e.g. canopy structure, subcanopy microclimate, soil properties, mowing regime) through management practices to enhance the colonization success of specialists in these linear habitats. These findings underpin the management needed to better conserving the biodiversity of agricultural landscapes across broad geographical scales.

Current analyses and predictions of spatially explicit patterns and processes in ecology most often rely on climate data interpolated from standardized weather stations. This interpolated climate data represents long-term average thermal conditions at coarse spatial resolutions only. Hence, many climate-forcing factors that operate at fine spatiotemporal resolutions are overlooked. This is particularly important in relation to effects of observation height (e.g. vegetation, snow and soil characteristics) and in habitats varying in their exposure to radiation, moisture and wind (e.g. topography, radiative forcing or cold-air pooling). Since organisms living close to the ground relate more strongly to these microclimatic conditions than to free-air temperatures, microclimatic ground and near-surface data are needed to provide realistic forecasts of the fate of such organisms under anthropogenic climate change, as well as of the functioning of the ecosystems they live in. To fill this critical gap, we highlight a call for temperature time series submissions to SoilTemp, a geospatial database initiative compiling soil and near-surface temperature data from all over the world. Currently, this database contains time series from 7,538 temperature sensors from 51 countries across all key biomes. The database will pave the way toward an improved global understanding of microclimate and bridge the gap between the available climate data and the climate at fine spatiotemporal resolutions relevant to most organisms and ecosystem processes.